How I ended up here, and what keeps me going.
Hi, I'm Adithya M D. I graduated with an engineering degree and absolutely no idea what to do with it. Most of my classmates picked corporate jobs. Sensible, stable, respectable choices. I, on the other hand, decided that solving messy real world problems sounded more interesting. Looking back, that was just a fancy way of saying I chose the harder path without fully understanding what I was getting into.
Early on, I worked with a few startups. Small teams, limited resources, lots of learning the hard way. Eventually, my friend Shreyas Kulur and I started something of our own. We ended up building a moisture meter for the cashew processing industry. That wasn't the original plan. We just kept meeting processors who were struggling with moisture control during processing, and we naively thought, "this can't be that hard to fix." It was. It took far longer and taught us far more than we expected. But over time, the device found its way into over 100 processing units across India and parts of Africa. Not through any big marketing push, just through word of mouth and people finding it genuinely useful. That still feels good to think about.
These days, I work as a Senior Project Associate at the Indian Institute of Science (IISc). My work sits at the intersection of embedded ML, audio and vision sensors, and building things that need to actually work outside of a lab. I've been lucky to collaborate with research teams and industries where the question isn't "does this work in theory?" but "does this survive a dusty factory floor?"
There's also a project I'm working on around insect monitoring and analysis that I'm quite excited about. I'll share more once we have proper field results. For now, it lives somewhere between "this is really promising" and "why on earth is the sensor doing that?"
I didn't get into machine learning because it was trendy or because someone told me it was the future. I got into it because real world data broke me.
When we were building the moisture meter, I spent a lot of time collecting data from factories, godowns, and dockyards. The data was noisy, inconsistent, and full of surprises. Rules I wrote by hand kept falling apart. Intuitions I trusted turned out to be wrong. That experience taught me something simple: when reality is messy, you need better tools to make sense of it. Data and mathematical reasoning became those tools for me.
That's what pulled me toward TinyML, signal processing, and edge inference. I'm fascinated by the challenge of making ML work on small, resource constrained hardware. No cloud, no GPU, just a tiny processor trying its best to be smart with very little memory.
This might sound strange, but I think teaching is one of the best ways to learn. Every time I've tried to explain something to a room full of people, I've discovered gaps in my own understanding that I didn't know existed. It keeps me honest.
I've been a Teaching Assistant for:
I should probably clarify: I have no plans of becoming a professor or climbing any academic ladder. I teach because I enjoy watching things click for people. There's a moment when someone goes from confused to "oh wait, I get it now" and that moment is genuinely one of my favourite things.
I'm still very much figuring things out. I build things, break them, learn something new, fix them, and then break them again in a different way. Everything on this site reflects my current understanding, which I fully expect to change as I keep going. If you come back in a few months, some of these pages might look completely different. That's kind of the point.